By Catarina Dutilh Novaes
This is the fourth and for now final installment of my series of posts on conceptual genealogy. Part I is here; Part II.1 is here; Part II.2 is here; a tentative abstract of 2 years ago, detailing the motivation for the project, is here.
I now discuss the five main features of the historicist conception of philosophical concepts that motivates and justifies the method of conceptual genealogy for philosophical concepts. In a sense, this is the backbone of the paper and of the whole project, so I'm particularly interested in feedback from readers now.
-----------------------------------------
We are now in a better position to describe in more detail what I take to be the five main characteristics of the historicist conception of philosophical concepts that I defend here, borrowing elements from Nietzsche’s conception of genealogy and Canguilhem’s concept-centered historical approach. In short, these are (they will each be discussed in turn subsequently):
- Historical change
- Superimposition of layers of meaning
- Contingency
- Multiple lines of influence
- Connected to (extra- or intra-philosophical) practices and goals
1. Historical change. This is probably the least controversial aspect of the historicist conception of philosophical concepts: few to none would deny that, as a matter of fact, philosophical concepts do change overtime. But the question then becomes: how philosophically relevant is it (as opposed to ‘merely historically’ relevant) to track these changes? As described in Part I (in connection with the work of Srinivasan), one possible response (and one that is often either explicitly given or presupposed) is that the temporal development of a philosophical concept X is not relevant for the determination of the ‘essence’ of X, i.e. the necessary and sufficient conditions for something to count as X. If past philosophers differed from us in their conceptualization of X, then either they were wrong or we are wrong (most likely the former…). We’d learn nothing of philosophical value about X by examining mistaken past conceptions of X.
According to the present approach, however, most (if not all) philosophical concepts are by and large theoretical constructs, even if they latch on in important ways to a concept-independent reality (just as Canguilhem’s concept of the ‘normal’ is closely related to the biological phenomenon of life, even if it undergoes modifications through time as a concept). Moreover, as they maintain traces of their previous instantiations throughout their development (as will be argued shortly), in particular in the form of presuppositions that then become uncritically accepted, to understand the current instantiation of a particular concept it becomes crucial to examine how it was construed over time. And thus, the historical development of philosophical concepts is not only a factual observation; it is also a philosophically relevant aspect for the analysis.
2. Superimposition of layers of meaning. The relevance of the philosophical history of a concept for the analysis of its current instantiations relies crucially on the idea of ‘superimposition of layers of meaning’, which we encountered when discussing Nietzsche’s conception of genealogy. An alternative conception would amount to, for example, viewing the temporal changes of a concept as corresponding to ‘radical revolutions’, after which nothing of their previous instantiations stays in place. If this were the case, then the analysis of past instantiations of a philosophical concept with the goal of improving our understanding of its current version(s) would be idle at best, but in fact also potentially distracting and misleading.
That the ‘superimposition of layers of meaning’ model offers a more accurate account of the development of philosophical concepts than the alternative ‘radical revolutions’ account is best argued for by means of detailed concrete examples (as will be done shortly). For now, let me offer a brief example to illustrate the point: the philosophical concept of substance received a number of formulations starting in ancient Greek philosophy, the most influential of which (historically speaking) was Aristotle’s conception. Through the centuries, it received a number of different interpretations, many of which radically different from the Aristotelian conception (e.g. Descartes’ notion of substance), while at the same time maintaining some key Aristotelian components such as the idea that there are basic elements that are the building blocks for all that exists in reality (Robinson 2014).
The main reason why it is of the utmost importance for the philosopher to be aware of the stratified nature of philosophical concepts is the fact that neglecting this dimension leads to the uncritical assimilation of presuppositions and substantive theoretical choices made along the way in the shaping of a concept, which then come to be viewed as truisms. These are often described as one’s absolute ‘intuitions’ about concept X, when in fact they are the products of theoretical choices made along the way. It is by exposing and investigating these different layers of meaning that the philosopher is able to isolate the theoretical choices that have led to the particular shaping of a given philosophical concept overtime.
Notice also that the ‘superimposition of layers’ model coupled with the idea of a stable core which fixes the object of the genealogy sits well with the idea of the polyvalence and plasticity of a (fruitful) philosophical concept, amenable to receiving a number of theoretical interpretations. Arguably, each of these interpretations will impose a new meaning to the concept in question, but traces of previous interpretations will remain.
3. Contingency. The process of superimposition of layers of meaning may be understood in at least two radically different ways: either in teleological or in contingent terms. On the teleological conception, each new instantiation of a concept is necessitated by its past instantiations: there is one unique path for the temporal development of a concept (roughly, in the spirit of Hegelian teleology).
On the contingentist conception, in contrast, each new meaning is one of the many new layers that a given concept is capable of acquiring at any given time, and a result of contingent events pertaining to the ‘struggle’ among the different ‘contenders’ (to resort again to Nietzschean ideas). In other words, the historical development of a concept branches to the future.
And thus, there are always possible theoretical paths that do not become actualized or instantiated for a given concept.[1] Which path does become instantiated (at the expense of the other possibilities) is typically a matter of substantive theoretical choices; however, once the choice is made, its substantive content often becomes viewed as a constitutive component of the concept as such, as if it had been there all along, and necessarily so. But of course, on the contingentist conception, this is a mistake: the historical development of concept X could have taken a rather different turn at some point or another. If this had been the case, then its current instantiations might have become something quite different. (Concrete examples will be discussed in the second half of the paper.)
One of the upshots of the contingencist conception is again that philosophical concepts are by and large theoretical constructs: they do not track necessary essences that exist in reality (though they do latch on to phenomena in reality), but rather unfold through time by means of the process of superimpositions of layers of meaning described above.
4. Multiple lines of influence. The contingency aspect just discussed ensures that the developmental path of a concept branches to the future. But what about the past? Well, as discussed when examining Nietzsche’s conception of genealogy, the instantiation of a concept or practice at a given point in time is typically the result of confluence of multiple lines of events and ideas, each having influenced the current status quo in different ways. And thus, the historical development of a concept branches to the past as well, not only to the future.
Both Leiter and Geuss note that, in this respect, Nietzschean genealogy differs from genealogy as pedigree, in that the latter countenances only one, non-branching line of ancestry leading to a given person: each person has one unique father. However, this is the case only if the sole lineage that matters in a genealogy is the male lineage. If the female lineages are included, then a person’s genealogical tree is composed of multiple lines of influence, namely the multiple branches and lineages that come together in one and the same person. Viewed from this perspective, the genealogy of a person is not that different from the genealogy of a concept or practice, as described by Nietzsche.
More concretely, this means that a philosophical concept will often be shaped by diverse lines of influence. One example, to be discussed in more detail below, is the concept of ‘logical form’; it emerged from the confluence of Aristotelian metaphysical hylomorphism with developments in logic, in particular Aristotelian syllogistic (Dutilh Novaes 2012). As noted by MacFarlane (2000), Aristotle is both the father of hylomorphism and the father of logic, but he is not the father of logical hylomorphism: he himself never brought the two concepts together. This was done only later, in the tradition of the ancient commentators.
5. Connected to (extra- or intra-philosophical) practices and goals.
The final component of the specific historicist conception of philosophical concepts presented here is the commitment to viewing such concepts not as disembodied constructions, floating above and beyond all human reality and human practices. Instead, philosophical concepts are typically embedded in (philosophical as well as non-philosophical) practices and goals, and are formulated as responses to specific theoretical needs of a given intellectual community.
The claim is not that philosophical concepts emerge only in response to material conditions and material needs, as the crude Marxist might claim (though these factors too are more significant than philosophers typically recognize). Rather, the idea is that philosophical concepts, theories and techniques typically emerge or receive new interpretations in response to theoretical or even practical needs, or against the background of pre-existing practices. For example, Aristotelian logic is best understood against the background of the practices of ancient dialectic, which we presume were pervasive in the early Academy (judging from e.g. Plato’s dialogues) (Kapp 1975). Ignoring the dialectical background causes much confusion concerning otherwise puzzling aspects of Aristotle’s syllogistic such as irreflexivity (Duncombe 2014).
An example closer to home would be the emergence of modern symbolic logic in the late 19th century. It is absolutely imperative to keep in mind the broader context for these developments, namely the projects of axiomatization of portions of mathematics. This was the primary goal and function of the logical systems designed by Frege, Russell, Hilbert etc. Understanding this background clarifies a number of still pervasive features in the practices of logicians, such as the prominent role of completeness proofs (Awodey & Reck 2002).
Another possibly informative example are medieval logical theories. When they began to be investigated more systematically in the mid-20th century, scholars tended to look at them only from the point of view of modern logic and modern concerns. Thus, Geach (1962) went on to claim that theories of supposition (a very important group of semantic theories in Latin medieval logic) were theories of reference – and what is more: they were very deficient theories of reference – without raising the question of whether theories of supposition might be theories of something else altogether. In my own work on theories of supposition, by contrast, the starting point was the question: why did the medieval authors themselves need something like a theory of supposition? What did they need these theories for? My answer to this question is that a semantic theory providing the tools for textual interpretation was much needed in a tradition where textual commentary and interpretation played such a crucial role. And so, medieval theories of supposition are best understood as theories of sentential meaning rather than as theories of reference (Dutilh Novaes 2008).
Thus the theoretical function(s) of a concept, broadly speaking, is always an important element in how a philosophical concept is shaped and interpreted at a given time. But this of course poses a serious problem for the historian/interpreter, who must inhibit the temptation of projecting her own ideas regarding the function(s) of a given concept, which presumably will reflect her own Zeitgeist, into past instantiations of the concept (as Geach seems to have done for supposition/reference). This is a point that Nietzsche was acutely aware of in his genealogical project, and he criticized his predecessors for failing to take it into account: one must avoid projecting current uses and meanings into the past.
On this point, Nietzsche follows Darwin, who cautioned against “the mistake of inferring current function or meaning from ancestral function or meaning” (Dennett 1995; 465; cf. Clark 1998a; xxiv). (Leiter 2015, 135)
In fact, mis-projection of meaning and function can take place in both directions: it is a mistake to infer past function or meaning from current ones, but it is also a mistake to infer current meaning and function from past ones. (Notice that this is a potential pitfall for ‘state-of-nature’ genealogical projects such as Craig’s (1990, 2007).) This is so because purpose and function are precisely elements of change and discontinuity in the historical development of a concept – given that theoretical needs and available techniques themselves change overtime.[2] And thus, the fact that a given concept has had a certain meaning in the past does not mean that the same meaning is the one that we, current philosophers, should attribute to it (see discussion of the genetic fallacy below).
However, the point is still that traces of these previous meanings (related to past purposes and functions) may still be present in the current embodiment of a concept, which in fact do not necessarily sit well with the current functions and purposes of a concept – and yet are uncritically accepted by current practitioners. Exposing such tensions is indeed one of the goals of conceptual genealogy, just as it is a goal for Nietzschean genealogy.
-----------------------------
So that's it for now, folks! I'll get back to working on this paper after the Xmas break. The remaining sections are:
II.4: Archeology or genealogy?
III. Applications
III.1: Genealogy as vindicating or debunking
III.2: Genealogy as expository (isolating assumptions)
III.3: The genetic fallacy
If there's enough interest, I'll post them here too in due course.
[1] In fact, different philosophical traditions may emerge precisely if two or more groups take different paths with respect to core philosophical concepts. So at a given point in time, it may well happen that more than one path is instantiated, but by different groups/traditions. This can be observed for example in the different interpretations given to some core Aristotelian concepts in the Latin and Arabic medieval traditions.
[2] A similar point holds of Canguilhem’s conception of the genealogy of concepts, which (on Méthot’s interpretation at least) is very much attuned to practical aspects such as the development of material techniques. According to Méthot, it is an ‘embodied-embedded’ conception of concepts, contrary to what some critics have maintained.
Recent Comments